Forearm surface electromyogram (sEMG) is a cumulative signal of the muscle contractions which leads to wrist and finger movement. In this study, we have collected six channels of sEMG from forearm muscles and touch forces at five fingertips simultaneously. Novel classification algorithms were then developed employing Time-Domain Auto-Regression (TD-AR), Principal Component Analysis (PCA), Artificial Neural Network (ANN), and Quadratic Models to predict finger motion and force from the forearm sEMG. The prediction models demonstrate high prediction accuracy and computational efficiency to estimate both the force and motion simultaneously. The models developed in this work provide a new approach for real time force and motion estimation in prosthetic limbs and other man-machine systems. © 2015 IEEE.